The gain is there before the powerful have decided how much of it to take. A thing is worth more to its buyer than it costs to provide; between those two amounts lies the reason to make the exchange, and across that interval the seller may move the price. This is the prize that gives market power its point. The firm seeks customers who cannot easily leave, a product whose substitutes disappoint, a permission its rivals cannot obtain. Each can help it bring the price nearer to what the buyer would have paid. But the means of reaching the prize matter. One seller takes a larger share of a trade that still occurs; another refuses a trade worth making; a third finances something that would otherwise never be supplied. The size of the receipt tells us too little until we know how it was earned.
A seller charging one price may gain by leaving some of the available surplus unmade. The reason is that the next sale can require a concession on all the previous ones. A merchant who cuts the price to attract a hesitant customer may have to extend that lower price to the eager customers already at his counter; their purchases become less profitable, and the new customer's business may not repay the sacrifice. The seller therefore compares his additional revenue with his additional cost, while the community's account compares the buyer's additional benefit with that cost. The two comparisons can recommend different quantities. The trades forgone at the seller's chosen price are part of the price of protecting his return on the trades he keeps.
Consider three customers, each wanting at most one identical object, valued by them at eighteen, twelve and six dollars. Suppose each object costs three dollars to supply, there are no fixed costs or effects on other people, and customers buy when the price equals their valuation. With competitive supply at three, all three buy and keep a combined twenty-seven dollars of surplus. A sole seller choosing one price can charge eighteen and earn fifteen, charge twelve and earn eighteen on two sales, or charge six and earn nine on three. Twelve is his best price. The two purchasers retain six dollars between them; the seller keeps eighteen; together they receive twenty-four. Compared with the competitive outcome, eighteen has moved to the seller and three has disappeared in the third customer's missing purchase. The seller's prize and the community's loss are different amounts, produced by different parts of the same decision.
The distinction survives a change in who receives the seller's money. Suppose a fixed franchise fee takes ten dollars of that eighteen and pays it to the town, with the business still operating and all other conditions unchanged. The best selling price remains twelve: subtracting the same ten from each operating choice leaves their ranking intact. The town receives ten, the owner retains eight, and the buyers retain six. The foregone gain remains three. A transfer of the prize has changed its beneficiary without repairing the exclusion that helped produce it. A charge on each unit sold would enter the quantity decision differently, while a fixed charge large enough to close the business could also change the result. The form of the claim matters as much as the name on the receipt.
A seller who can distinguish the customers may reach for the surplus more directly. Pigou separated the power to set a price from the further power to discriminate among purchasers or units. The distinction requires information and a means of maintaining different terms. A customer offered the expensive ticket must be unable simply to claim the cheap one, and a buyer receiving the cheap object must be unable profitably to resell it to someone facing a higher price. The obstacle may lie in the service itself, as when a treatment is supplied to a particular patient, or in the conditions governing access. Knowing that someone wants a thing dearly is useful to the seller only when that knowledge can be turned into terms the customer cannot evade.
Perfect discrimination makes the difference between capture and destruction particularly sharp. Give the seller in the example exact knowledge of the three valuations and the ability to make a separate offer to each, with resale prevented and acceptance at indifference retained. Prices of eighteen, twelve and six yield receipts of thirty-six against a total cost of nine. All three objects are supplied. The whole twenty-seven of available gain goes to the seller, and the buyers keep none. Output is efficient within the example, although the division is entirely one-sided. The seller has recovered the third transaction because its price no longer threatens the terms on which the first two are sold. A buyer's remaining surplus can fall while the total surplus rises; the two movements answer different questions.
The perfect case reveals a possibility, while ordinary discrimination leaves a more difficult account to settle. A membership fee, a premium version and a cheaper restricted version can each make some customers pay more without reproducing perfect knowledge of every valuation. Some lower prices admit buyers previously excluded; some conditions steer people toward products that cost more to make or serve their needs less well. Pigou's group-pricing analysis therefore distinguishes the aggregate quantity from its allocation across markets. Counting additional sales alone cannot settle the welfare question. One must ask what those buyers receive, which purchases elsewhere have disappeared, what supplying them costs, and how the gain is divided. The appetite for the prize is common to these arrangements; their consequences are not.
The prospect of capturing a surplus can also determine whether there will be anything to buy. Add to the three-customer example an avoidable setup cost of twenty-five dollars, incurred before any object can be supplied. The available gain after that cost is two. A seller restricted to one price can recover at most eighteen above variable costs and will not undertake the project. Under the perfect-discrimination assumptions, he can recover twenty-seven and retain two after paying for setup. The larger claim makes provision privately possible. This example gives a precise reason to distinguish a payment needed to bring a useful activity into existence from a payment extracted after that activity is already available. It gives no general license to charge any price: alter the cost, the alternatives or the terms of finance, and the necessity of the claim changes with them.
The relevant comparison must also allow for differences in production, rather than attribute every observed advantage to exclusion. Marshall asks whether concentrating supply can lower costs enough to change the comparison with a dispersed industry, and immediately makes ability, enterprise and access to capital part of that question. In our example, lowering the real cost of each object from three to two would enlarge the gains available from supplying all three from twenty-seven to thirty. Raising a price alone would not create those extra three dollars. A protected supplier may both improve production and charge restrictive terms; a competitive supplier may earn a return on an especially useful resource. The account must identify the productive change and the restriction separately before judging what the receipt represents.
The prize is therefore a useful way to read power without making every return tell the same story. Ownership, information and the ability to refuse access help determine who receives the gain. They may also affect its size, through exclusion, costly separation, investment or improvements in production. The seller's books combine these results in a single balance; the inquiry has to reopen that balance and follow the separate movements. How much would have existed under the available alternative? How much has merely changed hands? How much required the prospect of a return before anyone would bring it about? These questions carry the volume's argument from the size of the pie to the institutions that make a claim on it.
A larger prize can come from taking more, making more, or leaving less for anyone to take.
Power reaches into the interval between what people will pay and what provision costs, but that interval records both cooperation and the contest over its terms. A defensible account preserves both. It recognizes the gain a new service makes possible, follows the price that divides the gain, and counts the worthwhile exchanges a restriction prevents. The next claimant is the state, whose tax can change the division and the quantity at once. The Toll asks how to follow that claim past the person who hands over the money to the people whose alternatives bear its weight. The argument remains open.